BMC Oral Health

Table 1 Specificity, sensitivity, and accuracy of the deep learning models used in this study

From: Deep learning-based prediction of osseointegration for dental implant using plain radiography

	Specificity			Sensitivity			Accuracy
Model	Mean	Best	Worst	Mean	Best	Worst	Mean	Best	Worst
ResNet18	0.802	0.892	0.646	0.811	0.874	0.690	0.806	0.882	0.702
ResNet-34	0.810	0.933	0.687	0.832	0.895	0.748	0.822	0.886	0.759
ResNet-50	0.857	0.925	0.779	0.817	0.903	0.595	0.836	0.896	0.734
DenseNet-121	0.823	0.899	0.739	0.813	0.879	0.730	0.818	0.870	0.734
DenseNet-201	0.809	0.915	0.680	0.827	0.942	0.690	0.816	0.854	0.760
MobileNet-V2	0.816	0.883	0.705	0.833	0.922	0.659	0.824	0.883	0.755
MobileNet-V3	0.780	0.883	0.704	0.819	0.890	0.750	0.799	0.853	0.772

Back to article page

ISSN: 1472-6831

Contact us

Submission enquiries: bmcoralhealth@biomedcentral.com
General enquiries: ORSupport@springernature.com